We started out with organizations running their own infrastructure. Then with #AWS, #Azure, #GCloud, etc, the push was cloud infrastructure. It's distributed. It's more resilient. It can be distributed physically.

What mega corporation outages have taught me over time is that it's mostly smoke and mirrors. Shifted blame. One infra, even cloud, is a single point of failure.

Current example: #GitHub. All those resources and it still fails often. I'm far more comfortable running things myself.